Software for the process nobody sells

The work that is hardest to buy software for is usually the work that makes a business distinctive. It grew up over years, it fits the way this company does things, and nobody has ever written a package for it because there is no market of one.

So it ends up in a spreadsheet. Then a second spreadsheet. Then an email rule, a shared folder and a person who knows which column means what.

What that costs

Nothing, right up until it costs a great deal.

The file corrupts, or the person who understands it leaves, or the business grows enough that the manual steps stop being possible. Occasionally it is an auditor asking for a record that was never kept properly in the first place.

Until then the cost is quiet: time spent copying, errors found late, and decisions made on figures somebody assembled by hand and nobody wants to vouch for.

Some of what we have actually built

Not a list of sectors we target. A list of jobs that turned up, none of which had a package that fitted:

• A beam calculation program, now around thirty years old. The company that commissioned it was bought out, and three years later the new owner came back to us to move it onto a laptop with an automatic backup, because they still needed it
• Membership and class scheduling for a gymnastics club
• Survey and inspection records for an asbestos surveying firm
• A booking system for a small theatre
• Collection rounds for a waste oil business
• Order processing for a memorial masonry group, shipped in 1999 and still running

There are more, and the point is not the variety for its own sake. It is that a gymnastics club and an asbestos surveyor have nothing in common except the shape of the problem: a process the business depends on, with rules and exceptions nobody else has, held together by hand.

The beam calculator is worth dwelling on. When a business is acquired its software is normally the first thing to go, because the buyer has their own systems and every reason to consolidate. This one survived, and then three years later, as the company moved into brand new offices and a yard, they came back and asked us to carry it across onto a laptop and set up a backup that runs on its own.

A business that has just built new premises and could buy whatever it wanted chose to keep a thirty year old program. Nobody made that decision out of sentiment. They made it because the software still does the job better than the alternatives.

What we do

We build the thing the spreadsheet was standing in for. Entered once, checked automatically, recorded as it happens, and connected to the systems either side of it so nothing needs rekeying.

What the job is does not matter much. We have built stock control and order processing, production planning and CRM, and the ones that do not have a name are the same work: watch the process, understand the exceptions, build what matches.

And sometimes we say no

If the answer is a tidier spreadsheet, a change to the order of two steps, or a setting in software you already own, we will tell you that instead. It is a worse day for us and a better one for you, and it is the reason clients come back.

If you are weighing it up, when a spreadsheet stops being the right tool sets out the signs we look for.